def test_contour(grad): # FIXME: check the result m = Minuit( func3, grad=grad, pedantic=False, x=1.0, y=2.0, error_x=3.0, print_level=0 ) m.migrad() m.contour("x", "y")
def test_contour_with_gradient(): # FIXME: check the result m = Minuit(func3, grad_fcn=func3_grad, pedantic=False, x=1., y=2., error_x=3., print_level=0) m.migrad() m.contour('x', 'y')
# <codecell> #1D value Scan x,y = m.profile('x',subtract_min=True); plot(x,y) #if you have matplotlib # <codecell> #we also provide convenience wrapper for drawing it m.draw_profile('x'); # <codecell> #2d contour NOT minos contour x,y,z = m.contour('x','y',subtract_min=True) cs = contour(x,y,z) clabel(cs) # <codecell> #or a convenience wrapper m.draw_contour('x','z'); # <markdowncell> # ###Hesse and Minos # <markdowncell> # ####Hesse
def test_contour(): # FIXME: check the result m = Minuit(func3, pedantic=False, x=1., y=2., error_x=3., print_level=0) m.migrad() m.contour('x', 'y')
def test_contour_subtract(): m = Minuit(func0, x=1.0, y=2.0) m.migrad() m.contour("x", "y", subtract_min=True)
def test_contour(grad): # FIXME: check the result m = Minuit(func0, grad=grad, x=1.0, y=2.0) m.migrad() m.contour("x", "y")
# Minos calculates asymmetric errors for more complicated cases m.minos(var="slope") m.minos(var="offset") # Some output m.print_param() # And now contours of the fitted function # (very similar to testFit_V2.py version) fittedSlope = param[0]['value'] fittedOffset = param[1]['value'] fig2, ax2 = plt.subplots() # bound=4 --> got to +/- 4 sigma # bins=100 --> how many bins to do xx, yy, zz = m.contour('slope', 'offset', subtract_min=True, bound=4, bins=100) CS = ax2.contour(xx, yy, zz, [2.30, 5.99, 9.21]) fmt = {} strs = ['68%', '95%', '99%'] for l, s in zip(CS.levels, strs): fmt[l] = s ax2.clabel(CS, inline=True, fmt=fmt, fontsize=9) ax2.plot(fittedSlope, fittedOffset, 'ko') ax2.set_xlabel('slope') ax2.set_ylabel('offset') fig2.show() input('Enter something to continue ') # If we are particularly interested in one parameter # (say: the slope) we can scan the chisquared as a function # of "slope" minimizing wrt to the other parameter ("offset"